Arthropod Prey of Nestling Red-Cockaded Woodpeckers in the Upper Coastal Plain of South Carolina
Total Page：16
File Type：pdf, Size：1020Kb
Wilson Bull., 107(3), 1995, pp. 485495 ARTHROPOD PREY OF NESTLING RED-COCKADED WOODPECKERS IN THE UPPER COASTAL PLAIN OF SOUTH CAROLINA JAMES L. HANULA’ AND KATHLEEN E. FRANZREB* ABSTRACT.-FOLK nest cavities of the Red-cockaded Woodpecker (Picoides borealis) were monitored with automatic cameras to determine the prey selected to feed nestlings. Twelve adults were photographed making nearly 3000 nest visits. Prey in 28 arthropod taxa were recognizable in 65% of the photographic slides. Wood roaches in the genus (Parcoblutta) made up 69.4% of the prey fed to nestlings. Other common prey items were wood borer larvae (Cerambycidae or Buprestidae, 5.4%), Lepidoptera larvae (4.5%), spiders (Araneae, 3.6%), and ants (Formicidae, 3.1%). Wood roaches were the only prey items consistently taken by all four groups of birds; they made up 63.3 to 81.6% of the prey observed. Other common prey generally were taken in large numbers only by a single group of woodpeckers. During the breeding season these woodpeckers utilize relatively few common arthropods to feed nestlings. Received 10 Nov. 1994, accepted 20 Feb. 1995. The endangered Red-cockaded Woodpecker (Picoides borealis), found throughout the south from eastern Texas to the Atlantic coast, uses live pine tree boles and branches as primary foraging substrate (Hooper and Lennartz 1981, Porter and Labisky 1986), but little is known about their food habits. Determining Red-cockaded Woodpecker prey is critical to understanding the species’ foraging ecology. The purpose of this study was to develop an effective system for long-term monitoring of prey use by the Red-cockaded Woodpecker and to develop quantitative prey use data for this endangered species. We examined the diet of nestling Red- cockaded Woodpeckers as part of a larger study on the biology and ecol- ogy of insects associated with the bark of live pine trees. METHODS We conducted this study on the Savannah River Site (SRS), an 80,269 ha Dept. of Energy (DOE) nuclear production facility in Aiken County, South Carolina. The SRS is on the upper Atlantic Coastal Plain Physiographic Province. Its northern sector consists of uplands and sandhills where conditions are dry and the most common plant communities are longleaf pine (Pinus palustris) plantations and natural stands of longleaf pine and turkey oak (Quer- cus laevis). Loblolly pine (P. taeda) and bottomland hardwoods predominate on the more mesic areas and in riparian areas. Longleaf, loblolly, and slash (P. elliotii) pine stands occupy approximately 14,924 ha, 25,677 ha, and 12,011 ha, respectively (Knox and Sharitz 1990). Most of the stands are under 50 years of age. Potential Red-cockaded Woodpecker nesting ’ USDA Forest Service, Southeastern Forest Experiment Station, 320 Green St., Athens, Georgia 30602. 2044. 2 USDA Forest Service, Southeastern Forest Experiment Station, Dept. of Forest Resources, Clemson Univ., Clemson, South Carolina 29634.1003. 485 486 THE WILSON BULLETIN * Vol. 107, No. 3. September 1995 FIG. I. Nikon F4 camera with MF-24 multicontrol back and a Nikon SB-25 flash unit in a watertight housing constructed from an electrical box. The infrared sending unit (a) of a Trailmaster 1500 game monitor is attached to the cavity tree above the nest cavity (b) and the receiving unit (c) is attached at the base of the tree. The game monitor detects the presence of a bird at the cavity and trips the camera. habitat is sparse on the site because only 1521 ha are in pine stands 60 years old or older. The land surrounding the centrally located nuclear production areas has limited public access and is managed by the USDA Forest Service. It contains several Red-cockaded Woodpecker groups containing breeding pairs and one-to-several associated helpers. At the SRS, every woodpecker is banded with a U.S. Fish and Wildlife Service band and a unique set of plastic colored bands. We selected four of the groups at the SRS that had successfully bred during previous years (Franzreb. unpubl. data), and we placed our cameras near their nest cavity trees. Each of our camera systems (Fig. I) consisted of a Nikon F4 autofocus camera with a Sigma 5O@mm/f7.2 APO lens and a Nikon MF-24 multicontrol back. which held a tilm cassette allowing 250 exposures and imprinted the time and date of each exposure on the tilm. We attached a Nikon SB-25 autofocus Hash with an auxiliary battery pack to increase the number of Hashes between battery changes. We housed each camera in a watertight I5 X 38 X 43 cm tiherglass electrical box (Fig. I). We cut a 9.5 X 17.5 cm hole in one side of each box and fitted it with a high quality 4-mm-thick glass pane sealed to the box with silicone caulk and held in place with mirror mounting brackets. A second hole 2.5 cm diameter was drilled in the back of the box directly behind the camera viewfinder and stmilarly titted with glass. Styrofoam insulation ( I .3 cm thick) was attached to the inside of the lid, back. and two sides of the boxes to provide some protection from heat and to reduce camera noise. The camera and Hash were supported on separate metal brackets attached to 1.3 cm thick plywood which was fastened to the hack of the boxes. We mounted each camera unit on a 4 m tall Warren and Sweat Co. tripod style deer stand. A mounting and aiming bracket was constructed from 4 X 4 cm angle iron and 2- cm-thick plywood (Fig. I). This arrangement allowed us to aim the camera vertically by sliding the bottom camera housing attachment bolt through an arc cut in the plywood hack- ing of the aiming bracket. We assembled four of these camera systems and located them at Hunula and Frunzreb l PREY OF RED-COCKADED WOODPECKERS 487 Red-cockaded Woodpecker Clusters 2, 3, 18, 19, and 39 (numerical designations used by the USDA, Forest Service, Savannah River Forest Station). Because Group 2 nested early and Group 18 was late, we were able to use the four cameras at five nest cavities. We attached Trailmaster 1500 game monitors to the boles of the nest trees to trip the cameras (Fig. 1). The infrared sending units of the game monitors were 2-3 m above the nest cavities, and the receiving units were attached to the trees 1.5-2 m above the ground. The cameras were automatically tripped when adult birds returned to the nest cavity and interrupted the infrared beam produced by the game monitors. The Trailmaster units were factory-adjusted to have no limit on the number of photographs they would initiate. We set each unit to operate at its most sensitive level and to have a 30.set delay after a photo was taken to minimize multiple pictures of the same feeding and to eliminate photos of adults leaving the nest with fecal sacs. The Trailmasters were programmed to operated from 06: 00 to 20:30 h EST We used Ektachrome 200 slide film. The shutter speed was l/250 set, and the autoex- posure was set on shutter priority. The flash unit automatically metered existing light through the camera lens and adjusted itself for proper exposure. The rechargeable nickel cadmium batteries used in the flash units were replaced every day. The film was developed as uncut rolls, which allowed us to examine each frame in sequence. Each frame or slide was ex- amined with a stereoscope at 12X magnification, and arthropods were identified to the lowest taxonomic level possible. Time, date, and bird identification based on leg bands was re- corded for each slide. Arthropod sampling in the study area (Hanula and Franzreb. unpubl. data) and the arthropod collection at the Georgia Natural History Museum, Athens, Georgia, facilitated prey identification for this study. We examined the frequency of visits with different prey items for all groups combined, for each woodpecker group separately, and by the age of the nestlings. In addition, we compared the frequency of visits by breeding males, breeding females, and helper males. RESULTS The cameras recorded over 3000 nest visits, of which 2978 were of sufficient quality to be used in the analysis. Of the usable slides, 64.5% contained identifiable prey, and 20.9% contained prey items that were visible but could not be identified conclusively. In another 9%, the prey item was not visible, either because the bird’s head was turned away from the camera or because its head was already extended into the nest cavity. In 5.6% of the pictures, no food items were visible. We identified 28 arthropod taxa as prey from the photos (Table 1). Red-cockaded Woodpeckers feed their young either a large prey item or carry a number of small prey at one time. Only rarely did we see a photo of a bird bringing one small item, such as a single roach egg case or a small spider. We used separate listings to show whether the prey consisted of single or multiple items. For example, in Table 1 “insect larva” means that the birds were carrying a single large larva in each visit, while “insect larvae” means that they were carrying many small larvae. The predom- inant food items were wood roaches (Blattidae) in the genus Parcohlatta, which were present in 69.4% of the visits with recognizable prey. Other common prey items were wood-borer larvae (Cerambycidae or Bupres- 488 THE WILSON BULLETIN l Vol. 107, No. 3, September 1995 TABLE 1 TOTAL VISITS OF 12 ADULT RED-COCKADEDWOODPECKERS WITH ARTHROPODPREY TO FOUR NEST CAVITIES ON THE SAVANNAHRIVER SITE Number of Percentof Prey item observatmns VlSltS Wood roach 1310 69.4 (Parcoblatta sp.) Wood borer beetle (larva) 103 5.4 (Cerambycidae or Buprestidae) Moth (larva) 85 4.5 (Lepidoptera) Spider 68 3.6 (Araneae) Ant (larvae and adults) 58 3.1 (Formicidae)